An Alternating-Direction-Method of Multipliers-Incorporated Approach to Symmetric Non-Negative Latent Factor Analysis.

Large-scale undirected weighted networks are frequently encountered in big-data-related applications concerning interactions among a large unique set of entities. Such a network can be described by a Symmetric, High-Dimensional, and Incomplete (SHDI) matrix whose symmetry and incompleteness should be addressed with care. However, existing models fail in either correctly representing its symmetry or efficiently handling its incomplete data. For addressing this critical issue, this study proposes an Alternating-Direction-Method of Multipliers (ADMM)-based Symmetric Non-negative Latent Factor Analysis (ASNL) model. It adopts fourfold ideas: 1) implementing the data density-oriented modeling for efficiently representing an SHDI matrix's incomplete and imbalanced data; 2) separating the non-negative constraints from the decision parameters to avoid truncations during the training process; 3) incorporating the ADMM principle into its learning scheme for fast model convergence; and 4) parallelizing the training process with load balance considerations for high efficiency. Empirical studies on four SHDI matrices demonstrate that ASNL significantly outperforms several state-of-the-art models in both prediction accuracy for missing data of an SHDI and computational efficiency. It is a promising model for handling large-scale undirected networks raised in real applications.

STUDIES ON THE CONSTITUENTS OF ARTEMISIA ANNUA L / 药学学报

Six crystalline components were isolated from the lipophilic fraction of Artemisia annua L. They have been identified as four sesquiterpenes, one flavonol and one coumarin. Qinghaosu I and III are new sesquiterpenes. Five main constituents, camphene, iso-artemisia ketone, 1-camphor, β-carophyllene, and β-pinene were identified from the volatile oil of this herb.

Ultrathin and stable active layer of dense composite membrane enabled by poly(dopamine).

We demonstrate that dopamine is able to self-polymerize and adhere firmly onto the substrate, which can create a hierarchical structure comprising an ultrathin active layer and a porous support layer. Such an approach opens a novel way to fabricating highly efficient and stable composite materials including composite membranes. More specifically, in this study the composite membranes are fabricated by simply dipping microporous substrate in aqueous dopamine solution under mild conditions. Nanoindentation measurement reveals the tight adhesion of dopamine onto microporous substrate, which is ascribed to numerous pi-pi and hydrogen-bonding interactions. The chemical composition of the active layer is analyzed by XPS, which demonstrates the self-polymerization of dopamine. The water contact angle of the dopamine coated membranes is reduced remarkably compared with that of the uncoated counterpart. Stylus profiler measurements display that the poly(dopamine) thickness increases as the coating time increases. FESEM images of the membranes' cross section show that an active layer (<100 nm) is deposited on the porous polysulfone (PS) substrate. Positron annihilation spectroscopy (PAS) is introduced to probe the fractional free volume properties throughout the cross section of the composite membranes and reveal that after dopamine double-coating the active layer becomes thicker and more compact. Moreover, pH and concentration of the dopamine solution exert notable influence on the fractional free volume of the composite membranes. The as-prepared membranes are tentatively employed for pervaporative desulfurization and exhibits satisfying separation performance as well as durability. This facile, versatile, and efficient approach enables a promising prospect for the wide applications of such novel kinds of ultrathin composite materials.